{ "markdown": "

\n \"License\"\n \"Release\"\n \"Last\n \"Plugins\"\n

\n\n

NextEra Labs Claude Plugins

\n\nA curated marketplace of plugins for [Claude Code](https://docs.anthropic.com/en/docs/claude-code), extending it with development workflows, code review, and integrations.\n\n> **Important:** Review any plugin before installing. Plugins may include MCP servers, hooks, and scripts that execute on your machine. We cannot guarantee that third-party plugins will work as intended.\n\n---\n\n## Installation\n\n**Step 1 — Add the marketplace:**\n\n```\n/plugin marketplace add nexteralabs/ne-claude-plugins\n```\n\n**Step 2 — Install plugins:**\n\n```\n/plugin install codesmith@ne-claude-plugins\n/plugin install skillforge@ne-claude-plugins\n```\n\nThen `/reload-plugins` to apply.\n\n---\n\n# CodeSmith\n\nA development workflow that enforces spec refinement, TDD, and simplicity at every step.
\nFast code that breaks costs more than thoughtful code that ships clean.\n\n\n\n\n\n\n
\n\n### Spec first, code second\nNo ambiguity survives the design phase. Requirements are clarified, edge cases surfaced, and scope locked before a single line is written.\n\n### TDD by default\nWrite a failing test first. Then make it pass. Then refactor. No production code exists without a test that demanded it.\n\n\n\n### KISS over complexity\nThree clear lines beat a premature abstraction. No feature flags for hypotheticals. No helpers for one-time operations. Minimum code that works.\n\n### Multi-agent review\nSecurity, logic, and spec compliance reviewed by dedicated agents that catch what humans skip.\n\n
\n\n> The skill triggers automatically when you start any dev work. It drives the full lifecycle: understand the task, refine the spec, plan with TDD steps, implement test-first with subagents, review with parallel agents, and ship a clean PR. Jira integration is built in when the Atlassian MCP is available.\n\n---\n\n# SkillForge\n\nYour skills get smarter every time you run this.
\nPoint it at any skill, define what \"good output\" looks like, and it runs an autonomous experiment loop until the failure rate disappears.\n\n\n\n\n\n\n
\n\n### Parallel execution\nN runner agents spawn simultaneously — each gets the skill and a test input, executes it in isolation, returns clean output. A full 5-run experiment takes the same wall time as a single run.\n\n### Blind scoring\nA separate grader agent scores every output without knowing what mutation was attempted. No self-grading bias. Strict: if unsure, FAIL.\n\n\n\n### Checkpoint/resume\nArtifacts are saved after every experiment. If the loop gets interrupted, pick up exactly where you left off — no reruns, no lost progress.\n\n### Eval discriminability\nBefore the loop starts, SkillForge flags evals that trivially pass or fail 100% of the time. Prevents wasting 50 experiments optimizing against a broken signal.\n\n
\n\n> Say \"this skill is flaky\" or \"run skillforge on my X skill.\" SkillForge establishes a baseline, runs an autonomous mutation loop — one targeted change at a time — and delivers a terminal summary with the score improvement, top changes, and a merge prompt. The original SKILL.md is never touched until you approve.\n\n---\n\n## All Plugins\n\nDeveloped and maintained by [Nextera Labs](https://github.com/nexteralabs).\n\n| Plugin | Type | Description |\n|--------|------|-------------|\n| [**codesmith**](plugins/codesmith/) | skill, agents | Dev workflow with spec refinement, TDD, KISS, multi-agent review, ticket to PR |\n| [**skillforge**](plugins/skillforge/) | skill, agents | Autonomous skill optimizer — parallel execution, blind grading, checkpoint/resume |\n| [**drawio**](plugins/drawio/) | skill | Create and edit draw.io diagrams: flowcharts, architecture, sequence diagrams, and more |\n| [**code-audit**](plugins/code-audit/) | skill, agents | Multi-agent code review with parallel security and logic analysis |\n| [**secret-guard**](plugins/secret-guard/) | hooks | Pre-commit guard — scans staged files for leaked secrets before committing |\n| [**obsidian-vault**](plugins/obsidian-vault/) | mcp | Read and write notes in your Obsidian vault from Claude Code |\n| [**discord-notify**](plugins/discord-notify/) | mcp, commands | Send messages to Discord channels via bot API with guided setup |\n\n### skillforge\n\nAutonomously optimizes any Claude Code skill through a scored mutation loop. Spawns parallel runner agents to execute the skill, parallel grader agents to score outputs blindly, and keeps only mutations that improve the score. Supports checkpoint/resume, parallel candidate testing, and an eval discriminability check before the loop starts. Ends with a terminal summary and a prompt to merge the improved skill back into the original.\n\n```\n/plugin install skillforge@ne-claude-plugins\n\nuse: \"Init project\" prompt to setup a project\n```\n\n### code-audit\n\nAsk Claude to review your code, a PR, or your current changes. The skill dispatches parallel security and logic review agents, then aggregates findings with confidence scores: critical issues, suggestions, and nits. No commands to remember, just say \"review my code\" or \"check this PR.\"\n\n```\n/plugin install code-audit@ne-claude-plugins\n```\n\n### secret-guard\n\nPre-commit hook that scans `git diff --cached` for API keys, tokens, passwords, private keys, and cloud credentials. Blocks the commit if anything is found. Zero overhead on all other commands.\n\n```\n/plugin install secret-guard@ne-claude-plugins\n```\n\n### obsidian-vault\n\nMCP server connecting Claude Code to your Obsidian vault. Requires `OBSIDIAN_VAULT_PATH` environment variable.\n\n```\n/plugin install obsidian-vault@ne-claude-plugins\n```\n\n### discord-notify\n\nSend messages to Discord channels. The skill detects when you need to notify a channel and handles it. Just say \"send a message to #general.\" Requires `~/.claude/discord.env` with bot token and guild ID (run `/discord-setup` for guided configuration).\n\n```\n/plugin install discord-notify@ne-claude-plugins\n```\n\n### drawio\n\nAsk Claude to create any kind of diagram: flowcharts, architecture diagrams, sequence diagrams, swimlanes, ER diagrams, network maps. The skill triggers whenever you want to visualize something and generates `.drawio` XML files you can open in draw.io or VS Code.\n\n```\n/plugin install drawio@ne-claude-plugins\n```\n\n---\n\n
\nContributing\n\nWe welcome contributions! See [CONTRIBUTING.md](.github/CONTRIBUTING.md) for the full guide.\n\n**Quick start:**\n\n1. Fork the repo\n2. Add your plugin under `plugins/`\n3. Register it in `.claude-plugin/marketplace.json`\n4. Validate frontmatter: `bun .github/scripts/validate-frontmatter.ts`\n5. Open a pull request\n\nYou can also submit a plugin via the [Plugin Submission](https://github.com/nexteralabs/ne-claude-plugins/issues/new?template=plugin-submission.yml) issue template.\n\n
\n\n
\nDevelopment\n\n### Validation\n\nAll skills and agents are validated for correct YAML frontmatter on every PR:\n\n```bash\nbun .github/scripts/validate-frontmatter.ts\n```\n\n### Testing Locally\n\nTest a plugin in development:\n\n```bash\nclaude --plugin-dir /path/to/your-plugin\n```\n\n### Repository Structure\n\n```\nne-claude-plugins/\n├── .claude-plugin/\n│ └── marketplace.json # Plugin registry\n├── plugins/ # Internal plugins (Nextera Labs)\n└── .github/\n ├── scripts/ # Validation tooling\n └── workflows/ # CI automation\n```\n\nFor more information on developing Claude Code plugins, see the [official documentation](https://docs.anthropic.com/en/docs/claude-code/plugins).\n\n
\n\n
\nLicense\n\nThis project is licensed under the [MIT License](LICENSE).\n\nCopyright 2026 Nextera Labs.\n\n
\n", "bytes": 8338, "sha": "44bea7d3242bb73bfa097530f12a2e2ee07bfad76a01b457b0db1b7385bb8c2f", "repo_slug": "nexteralabs/ne-claude-plugins", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/plg_nexteralabs_ne_claude_plugins_codesmith_6e8d9a46/readme" }